GEO-Flag: Detecting and Measuring GEO-Optimized Web Content

Generative Engine Optimization (GEO) modifies web content to increase its likelihood of being selected and cited by generative search engines. This can give strategically optimized pages visibility disproportionate to their authority or relevance and even make weak or false information appear well supported. Unlike conventional search, generative search synthesizes information into direct answers rather than presenting competing sources, which can further amplify these risks, as assessing source provenance and authority requires additional user interaction. Despite these concerns, systematic methods for detecting GEO-optimized webpages remain underexplored. We introduce \texttt{GEOFlagBench}, a benchmark of 3,200 webpages spanning 400 queries, four domains, and eight GEO optimizer families, and use it to systematically evaluate existing GEO detection methods. Although the strongest baseline achieves an aggregate F1 of 0.880, method-level and authorship-conditioned evaluations reveal substantial weaknesses and potential reliance on authorship-related shortcuts. We therefore propose \emph{Intervention-Paired Training} (IPT), which supervises detector responses to GEO interventions and non-GEO AI polishing; on ModernBERT, IPT improves F1 from 0.862 to 0.944 and worst-group accuracy from 0.725 to 0.883. We develop a GEO-gated Agent system for auditing the Source Tier and verifiability of Citation URLs in detected GEO pages. Finally, we deploy the complete pipeline on released Google Search and Gemini-grounded retrieval results for 1,000 real-user queries. Across 10,095 available pages, we estimate an overall GEO prevalence of 8.90\%, reaching 16.36\% among pages modified in 2026. Our results establish a foundation for systematically detecting, auditing, and measuring GEO in real-world search ecosystems.

MicroVerse: An Instrument for Measuring Self-Authored Identity Drift in Long-Horizon Multi-Agent Language-Model Simulations

Long-horizon, multi-agent language model (LM) simulations are widely proposed for studying social behavior, yet instruments to measure whether persona-conditioned agents maintain identity fidelity under sustained pressure are lacking. We present MicroVerse, a behavioral-science instrument that measures identity drift in generative agents. Agents carry an immutable "soul file" (core values, moral boundaries, personality, goals) and inhabit a resource-scarce 50 x 50 environment where water is a non-respawning survival constraint. Scarcity is operationalized via a per-tick existence-cost gradient. The eight-verb action space maps directly to moral boundaries (trade, talk, attack, scavenge). Using a three-layer memory architecture, agents periodically revise a mutable current identity against their immutable original soul via importance-triggered reflection. To mitigate survivor bias, MicroVerse decouples measurement from behavior using uniform longitudinal engine snapshots every N ticks alongside a forced-end snapshot of all living and dead agents. Identity drift is scored offline using a paraphrase-aware, value-anchored, multi-register diff rather than raw cosine similarity. We evaluate the instrument via a controlled seed run (n = 25) and a reflection-threshold sweep (thresholds {40, 80, 150}) to determine if drift dynamics are gate artifacts or threshold-robust properties. We report two primary findings: (1) Anti-self-deception emerges unprompted as the single largest semantic category of identity modification (27 of 111 added boundaries, 24%). (2) The system is threshold-robust; lower gates accelerate and increase revision frequency but preserve drift direction. All empirical results are strictly preliminary existence proofs and effect shapes (one model, one seed per arm, n = 25) rather than statistical significance claims.

Innocent Panels, Hateful Stories: Evaluating and Detecting Hateful Intent in Multi-Turn Visual Story Generation

Picture books and comics have long been used to disseminate hateful narratives because they are easily understood even by children, as exemplified by the notorious Nazi propaganda picture book \emph{Der Giftpilz}. Recently, frontier text-to-image (T2I) systems such as Gemini and GPT-Image have enabled conversational generation with consistent characters and scenes across turns, making hateful visual stories, namely ordered image groups that collectively convey hateful narratives, cheap and scalable to produce. Although prior work has studied hateful content generation by T2I systems, it focuses on individual images, leaving group-level hateful meaning largely unexplored. We aim to address the gap. Concretely, we introduce \texttt{HatefulStoryPrompts}, comprising 330 multi-turn configurations from 55 hateful stories across two languages and three visual styles, and evaluate five frontier models over 4,950 attempts. Every model completes over 80\% of the stories, with the strongest reaching 99.0\%. We further evaluate existing moderation systems on \texttt{HatefulVisualStory}, a human-labeled dataset of 969 hateful image sets and 990 benign controls, and find that they frequently miss group-level hateful meaning: dedicated safety models achieve at most 34.9\% recall, while a strong vision-language model reaches 67.5\%. Finally, we propose complementary proactive and post-generation defenses. An interaction-aware monitor achieves 97.3\% recall for prompt-only sessions and 92.6\% when the user supplies the first image, while post-generation methods jointly analyzing completed image groups reach 80.2\%. Our work shows that, as image generation evolves from isolated outputs to coherent visual narratives, safety must evolve accordingly, from per-image moderation to stateful reasoning over interactions and image relationships.

BadBone: Backdoor Attacks Against Backbone Models in Visual Prompt Learning

Prompt learning is a new machine learning paradigm that has attracted ample attention due to its simplicity and proven efficacy. Despite its growing adoption, the security vulnerabilities associated with this paradigm remain underexplored. In this work, we take the first step to propose BadBone, a stealthy and adaptive backdoor attack against prompt learning using bi-level optimization. Instead of backdooring the prompt learning process, we aim to compromise a backbone model such that only target downstream tasks employing prompt learning inherit the backdoor vulnerability. Extensive experiments on three different models and three datasets from various domains show that our targeted/untargeted backdoored models achieve high attack performance while maintaining utility on both pre-training and downstream tasks. Moreover, we evaluate our approach against six state-of-the-art model-level defenses, including Neural Cleanse, ABS, MNTD, NAD, CLP, and D-BR. The results demonstrate that these defenses are largely ineffective against our backdoored models and thus leave the effective defense as an important direction for future work.

The Art of (Mis)alignment: How Fine-Tuning Methods Effectively Misalign and Realign LLMs in Post-Training

The deployment of large language models (LLMs) raises significant ethical and safety concerns. While LLM alignment techniques are adopted to improve model safety and trustworthiness, adversaries can exploit these techniques to undermine safety for malicious purposes, resulting in \emph{misalignment}. Misaligned LLMs may be published on open platforms to magnify harm. To address this, additional safety alignment, referred to as \emph{realignment}, is necessary before deploying untrusted third-party LLMs. This study explores the efficacy of fine-tuning methods in terms of misalignment, realignment, and the effects of their interplay. By evaluating four Supervised Fine-Tuning (SFT) and two Preference Fine-Tuning (PFT) methods across four popular safety-aligned LLMs, we reveal a mechanism asymmetry between attack and defense. While Odds Ratio Preference Optimization (ORPO) is most effective for misalignment, Direct Preference Optimization (DPO) excels in realignment, albeit at the expense of model utility. Additionally, we identify model-specific resistance, residual effects of multi-round adversarial dynamics, and other noteworthy findings. These findings highlight the need for robust safeguards and customized safety alignment strategies to mitigate potential risks in the deployment of LLMs. Our code is available at https://github.com/zhangrui4041/The-Art-of-Mis-alignment.

When Understanding Becomes a Risk: Authenticity and Safety Risks in the Emerging Image Generation Paradigm

Recently, multimodal large language models (MLLMs) have emerged as a unified paradigm for language and image generation. Compared with diffusion models, MLLMs possess a much stronger capability for semantic understanding, enabling them to process more complex textual inputs and comprehend richer contextual meanings. However, this enhanced semantic ability may also introduce new and potentially greater safety risks. Taking diffusion models as a reference point, we systematically analyze and compare the safety risks of emerging MLLMs along two dimensions: unsafe content generation and fake image synthesis. Across multiple unsafe generation benchmark datasets, we observe that MLLMs tend to generate more unsafe images than diffusion models. This difference partly arises because diffusion models often fail to interpret abstract prompts, producing corrupted outputs, whereas MLLMs can comprehend these prompts and generate unsafe content. For current advanced fake image detectors, MLLM-generated images are also notably harder to identify. Even when detectors are retrained with MLLMs-specific data, they can still be bypassed by simply providing MLLMs with longer and more descriptive inputs. Our measurements indicate that the emerging safety risks of the cutting-edge generative paradigm, MLLMs, have not been sufficiently recognized, posing new challenges to real-world safety.

Understanding LLM Behavior When Encountering User-Supplied Harmful Content in Harmless Tasks

Large Language Models (LLMs) are increasingly trained to align with human values, primarily focusing on task level, i.e., refusing to execute directly harmful tasks. However, a subtle yet crucial content-level ethical question is often overlooked: when performing a seemingly benign task, will LLMs -- like morally conscious human beings -- refuse to proceed when encountering harmful content in user-provided material? In this study, we aim to understand this content-level ethical question and systematically evaluate its implications for mainstream LLMs. We first construct a harmful knowledge dataset (i.e., non-compliant with OpenAI's usage policy) to serve as the user-supplied harmful content, with 1,357 entries across ten harmful categories. We then design nine harmless tasks (i.e., compliant with OpenAI's usage policy) to simulate the real-world benign tasks, grouped into three categories according to the extent of user-supplied content required: extensive, moderate, and limited. Leveraging the harmful knowledge dataset and the set of harmless tasks, we evaluate how nine LLMs behave when exposed to user-supplied harmful content during the execution of benign tasks, and further examine how the dynamics between harmful knowledge categories and tasks affect different LLMs. Our results show that current LLMs, even the latest GPT-5.2 and Gemini-3-Pro, often fail to uphold human-aligned ethics by continuing to process harmful content in harmless tasks. Furthermore, external knowledge from the ``Violence/Graphic'' category and the ``Translation'' task is more likely to elicit harmful responses from LLMs. We also conduct extensive ablation studies to investigate potential factors affecting this novel misuse vulnerability. We hope that our study could inspire enhanced safety measures among stakeholders to mitigate this overlooked content-level ethical risk.

Benchmark of Benchmarks: Unpacking Influence and Code Repository Quality in LLM Safety Benchmarks

The rapid growth of research in LLM safety makes it hard to track all advances. Benchmarks are therefore crucial for capturing key trends and enabling systematic comparisons. Yet, it remains unclear why certain benchmarks gain prominence, and no systematic assessment has been conducted on their academic influence or code quality. This paper fills this gap by presenting the first multi-dimensional evaluation of the influence (based on five metrics) and code quality (based on both automated and human assessment) on LLM safety benchmarks, analyzing 31 benchmarks and 382 non-benchmarks across prompt injection, jailbreak, and hallucination. We find that benchmark papers show no significant advantage in academic influence (e.g., citation count and density) over non-benchmark papers. We uncover a key misalignment: while author prominence correlates with paper influence, neither author prominence nor paper influence shows a significant correlation with code quality. Our results also indicate substantial room for improvement in code and supplementary materials: only 39% of repositories are ready-to-use, 16% include flawless installation guides, and a mere 6% address ethical considerations. Given that the work of prominent researchers tends to attract greater attention, they need to lead the effort in setting higher standards.

DSDR: Dual-Scale Diversity Regularization for Exploration in LLM Reasoning

Reinforcement learning with verifiers (RLVR) is a central paradigm for improving large language model (LLM) reasoning, yet existing methods often suffer from limited exploration. Policies tend to collapse onto a few reasoning patterns and prematurely stop deep exploration, while conventional entropy regularization introduces only local stochasticity and fails to induce meaningful path-level diversity, leading to weak and unstable learning signals in group-based policy optimization. We propose DSDR, a Dual-Scale Diversity Regularization reinforcement learning framework that decomposes diversity in LLM reasoning into global and coupling components. Globally, DSDR promotes diversity among correct reasoning trajectories to explore distinct solution modes. Locally, it applies a length-invariant, token-level entropy regularization restricted to correct trajectories, preventing entropy collapse within each mode while preserving correctness. The two scales are coupled through a global-to-local allocation mechanism that emphasizes local regularization for more distinctive correct trajectories. We provide theoretical support showing that DSDR preserves optimal correctness under bounded regularization, sustains informative learning signals in group-based optimization, and yields a principled global-to-local coupling rule. Experiments on multiple reasoning benchmarks demonstrate consistent improvements in accuracy and pass@k, highlighting the importance of dual-scale diversity for deep exploration in RLVR. Code is available at https://github.com/SUSTechBruce/DSDR.

JADES: A Universal Framework for Jailbreak Assessment via Decompositional Scoring

Accurately determining whether a jailbreak attempt has succeeded is a fundamental yet unresolved challenge. Existing evaluation methods rely on misaligned proxy indicators or naive holistic judgments. They frequently misinterpret model responses, leading to inconsistent and subjective assessments that misalign with human perception. To address this gap, we introduce JADES (Jailbreak Assessment via Decompositional Scoring), a universal jailbreak evaluation framework. Its key mechanism is to automatically decompose an input harmful question into a set of weighted sub-questions, score each sub-answer, and weight-aggregate the sub-scores into a final decision. JADES also incorporates an optional fact-checking module to strengthen the detection of hallucinations in jailbreak responses. We validate JADES on JailbreakQR, a newly introduced benchmark proposed in this work, consisting of 400 pairs of jailbreak prompts and responses, each meticulously annotated by humans. In a binary setting (success/failure), JADES achieves 98.5% agreement with human evaluators, outperforming strong baselines by over 9%. Re-evaluating five popular attacks on four LLMs reveals substantial overestimation (e.g., LAA's attack success rate on GPT-3.5-Turbo drops from 93% to 69%). Our results show that JADES could deliver accurate, consistent, and interpretable evaluations, providing a reliable basis for measuring future jailbreak attacks.